A Feature Weight Adjustment Algorithm for Document Categorization

In recent years we have seen a tremendous growth in the volume of text documents available on the Internet, digital libraries, news sources, and company-wide intra-nets. Automatic text categorization, which is the task of assigning text documents to pre-speci ed classes (topics or themes) of documents, is an important task that can help both in organizing as well as in nding information on these huge resources. In this paper we present a fast iterative feature weight adjustment algorithm for the linear-complexity centroid based classi cation algorithm. Our algorithm uses a measure of the discriminating power of each term to gradually adjust the weights of all features concurrently. We experimentally evaluate our algorithm on the Reuters-21578 and OHSUMED document collection and compare it against a variety of other categorization algorithms. Our experiments show that feature weight adjustment improves the performance of the centroid-based classi er by 2%{5% , substantially outperforms Rocchio andWidrow-Ho and is competitive with SVM.